Yeast forms dominate fungal diversity in the deep oceans

Abstract

Fungi are the principal degraders of biomass in most terrestrial ecosystems. In contrast to surface environments, deep-sea environmental gene libraries have suggested that fungi are rare and non-diverse in high-pressure marine environments. Here, we report the diversity of fungi from 11 deep-sea samples from around the world representing depths from 1,500 to 4,000 m (146-388 atm) and two shallower water column samples (250 and 500m). We sequenced 239 clones from 10 fungal-specific 18S rRNA gene libraries constructed from these samples, from which we detected only 18 fungal 18S-types in deep-sea samples. Our phylogenetic analyses show that a total of only 32 fungal 18S-types have so far been recovered from deep-sea habitats, and our results suggest that fungi, in general, are relatively rare in the deep-sea habitats we sampled. The fungal diversity detected suggests that deep-sea environments host an evolutionarily diverse array of fungi dominated by groups of distantly related yeasts, although four putative filamentous fungal 18S-types were detected. The majority of our new sequences branch close to known fungi found in surface environments. This pattern contradicts the proposal that deep-sea and hydrothermal vent habitats represent ancient ecosystems, and demonstrates a history of frequent dispersal between terrestrial and deep-sea habitats.

Figure 1

Phylogeny demonstrating the branching position of deep-sea SSU types with a wide diversity of opisthokont sequences and sequences from general eukaryotic SSU eDNA surveys. A subsection of the fungal SSU rDNA phylogeny is shown, including other opisthokont groups such as chytrids, Zygomycetes and Glomeromycetes. Tree topology shown was calculated using PHYML (Guindon et al. 2005) from an alignment of 220 sequences and 1098 characters. Nodes in the phylogeny supported by in excess of 95% bootstrap support and 0.95 posterior probabilities in all the three analysis methods are represented by a black circle on the relevant node. Topology support values are given in the following order: Bayesian posterior probabilities calculated using MrBayes/100 PHYML (Guindon et al. 2005) bootstraps and 1000 LogDet distance bootstraps, and are shown when in excess of 0.75 and 50%, respectively. Non-fungal opisthokonts (Adl et al. 2005) are marked with a grey box bar. Higher fungal taxonomic groups labelled, according to Adl et al. (2005) or as listed in GenBank taxonomy databases, are marked with a shaded grey bar. Highly novel sequences potentially representing highly divergent lineages or unidentified taxonomic groups are placed within green boxes. All sequences are listed with GenBank accession numbers and environmental gene library sequences are marked according to crude environmental type listed in the key.

Figure 2

Subsection of the fungal SSU rDNA focusing on the basidiomycete section of the phylogeny. See figure legend 1 for details on diagrammatic conventions. A wide diversity of basidiomycete fungi detected in deep-sea environments are shown. The diversity detected tends to branch very close to basidiomycete fungi with an yeast growth form with notable exceptions, including a close deep-sea relative of the fruiting bodied brown-rot fungi Antrodia.

Figure 3

Subsection of the fungal SSU rDNA focusing on the ascomycete section of the phylogeny. See figure legend 1 for details on diagrammatic conventions. Similar to the basidiomycete analyses (figure 2), a wide diversity of ascomycete fungi detected in deep-sea environments are shown, with the deep-sea ascomycete fungi detected generally grouping with microbes with an yeast growth form. However, again, we see exceptions to this, including the detection of a deep-sea Cladosporium and Aspergillus SSU type.

Figure 4

Phylogenies of fungal sequences generated by this study in the context of their closest BLASTn matches. Boxed sequences are unique 18S-types (as defined by the fragment analysed: V4–V5 of 18S rDNA) detected in this study from deep-sea and marine environments. Phylogenies were calculated from highly similar sequences, therefore parsimony methods were used with gaps treated as a fifth character state. One thousand bootstrap replicates were calculated using the same parsimony methods and branches with 95% support are represented by a black circle on the relevant node. All trees are shown unrooted. Grey lines are only for labelling purposes. The scale bar for each phylogeny denotes the number of changes across the given branch length. The number of positions used to calculate each phylogeny is given in brackets below the scale bar.

Figure 5

Sampling saturation analyses of our marine fungal molecular survey demonstrate that the sampling of our deep-sea libraries was close to saturation. Species accumulation curve and cumulative Chao1 total diversity estimate for all deep-sea libraries are plotted.